Conversation
Replaces print statements with logger calls for better logging consistency across modules. Updates docstrings and comments to clarify the distinction between tool-level and retrieval-level intent classification, especially in intent_classifier.py and mcp_router/intent.py. Improves TOON format handling to preserve structured results for internal callers while optimizing external output. Adds debug and composability improvements to hybrid search and context modules.
The 'symbol' field is now included in compact TOON-formatted outputs in both search and context results. The TOON helper is now imported from a shared location to avoid duplication, with a fallback for standalone usage. Test logging for TOON decode errors now uses the logging module instead of print to prevent stdio issues.
Introduces relative path extraction and indexing in the Neo4j backend for improved filtering and cleaner output, updating all relevant Cypher queries and node properties. Enhances fallback cache implementations in hybrid_search.py and semantic_expansion.py with bounded OrderedDicts for FIFO eviction. Upgrades mcp to 1.25.0. Extends VICReg with batch entropy regularization (VICReg-E) to improve activation diversity and prevent collapse, with validated loss improvements.
🤖 Augment PR SummarySummary: This PR refines retrieval/search behavior and improves observability while preserving JSON-RPC compatibility. Changes:
Technical Notes: These changes primarily target better token efficiency, safer logging under JSON-RPC, and clearer intent routing boundaries. 🤖 Was this summary useful? React with 👍 or 👎 |
![augmentcode[bot]](https://avatars.githubusercontent.com/in/1027498?s=60&v=4){kind=small}
| if isinstance(results, list): | ||
| # Replace JSON array with TOON string (handles empty arrays too) | ||
| # Preserve original list for internal callers before TOON encoding | ||
| response["results_json"] = results |
There was a problem hiding this comment.
Including the full structured results_json alongside TOON-encoded results can negate the token-saving goal and may cause unexpectedly large responses for external clients. Consider whether results_json should be returned only in internal/composability contexts rather than in the tool’s public response.
Other Locations
scripts/mcp_impl/toon.py:116
🤖 Was this useful? React with 👍 or 👎
Introduces Intent-Aware Expansion Cascade (IAEC) for routing query expansion strategies based on query intent, adds pattern-based expansion using discovered code patterns, and implements calibrated confidence for reranking with score separation-aware early stopping. Updates configuration documentation and .env.example with new options. Refactors hybrid expansion and search to support intent-driven strategies, and enhances pipeline edge extraction for Neo4j/Qdrant compatibility. (This fixes QDRANT graph only collections on K8s deploy
Added upload-pvc.yaml, code-models-pvc.yaml, upload-service.yaml, and learning-reranker-worker.yaml to the Kubernetes deployment process and kustomization.yaml. Updated Dockerfile to include templates directory. These changes support new service deployments and resource management.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
No description provided.